Here to help
Unsure what you need? Got technical questions? Our team are here to help you through the whole process, from design to delivery.
GPU Optimised Solutions
As a NVIDIA Preferred Solutions Provider Novatech are trusted for our expertise in high-performance parallel computing by some of the UKs leading names in the simulation and life sciences industries.
Our Server solutions support the latest NVIDIA TESLA GPU architecture and are designed and engineered to deliver faster more efficient performance to accelerate your workflow.
Whether it's for Deep learning and AI accelerated analytics, or the ability to quickly process large data sets, GPU-accelerated computing offers an unprecedented performance boost by offloading compute-intensive tasks of the application from the CPU to the GPU. While a CPU consists of a few cores optimised for sequential serial processing, the GPU has a massively parallel architecture consisting of thousands of smaller, more efficient cores designed for handling multiple tasks simultaneously.
This results in up to 30 times the performance of CPU-only solutions, enabling the deployment of smaller, more efficient and computationally dense systems.
As an NVIDIA preferred Solutions Provider, Novatech work closely with NVIDIA to combine our technology expertise and services to design, supply, and implement advanced NVIDIA computing solutions.
Fully customisable
We customise all of our systems to match your budgetary needs and performance requirements. We don't do 'one size fits all'.
Tried & trusted
We've been in business since 1987 and are trusted by some of the leading names in the Aviation, Defence and Marine industries.
SuperMicro Dual Xeon GPU optimised Server
- Processor:Supports Dual Xeon E5-2600V4
- GPU Support:Up to 3x Double width GPU cards
- Memory:Supports up to 2TB ECC 3DS LRDIMM
- Drive Bays:4x 2.5" SATA/SAS Hot Swap bays
- Features:Redundant 1600W 80plus 95% (Platinum) high efficiency PSU.
1U
SuperMicro AMD Opteron GPU optimised Server
- Processor:Supports ultra high speed AMD Opteron™ 6000 Series
- GPU Support:Up to 2x PCI-E 2.0 x16 compatible NVIDIA Tesla cards
- Memory:Supports up to 512GB DDR3
- Drive Bays:3x 3.5" Enterprise SATA Hot Swap bays
- Features:1400W high-efficiency power supply with PMBus
1U
SuperMicro Dual Xeon High Density GPU optimised Server
- Processor:Supports Dual Xeon E5-2600V4
- GPU Support:Up to 4x mid mounted Tesla GPUS or Xeon Phi
- Memory:Supports up to 2TB ECC 3DS LRDIMM or 1TB ECC RDIMM
- Drive Bays:10x 2.5" Hot-swap drive bays
- Features:Three further rear PCI-e expansion slots available, Dual Redundant 2000W (80plus Platinum 94%) PSU system
2U
SuperMicro Dual Xeon 10-GPU with Single Root Complex Server
- Processor:Supports Dual Xeon E5-2600V4
- GPU Support:Up up to 10 Double-Width NVIDIA Tesla GPUS
- Memory:Supports up to 3TB ECC 3DS LRDIMM or 1.5TB ECC RDIMM
- Drive Bays:24x 2.5" bays
- Features:Supports GPUs under Single-Root complex RDMA support | 2000W Redundant Power Supplies with PMBus
4U
SuperMicro Dual Xeon Rack/Tower GPU optimised Server
- Processor:Supports Dual Xeon E5-2600V4
- GPU Support:Up to 4x double-width NVIDIA Tesla or Xeon Phi
- Memory:Supports up to 2TB ECC 3DS LRDIMM or 1TB ECC RDIMM
- Drive Bays:8x 3.5" Hot-swap SAS/SATA
- Features:2000W Redundant Power Supplies with PMBus | Rackmountable / Tower
4U
Can’t find what you’re looking for?
The systems listed on this page are just a small selection of what we can offer. So if you’re looking for something in particular, Contact us and we'll help you design a custom solution to meet your needs.
Our expertise
How we work
We are an open-architecture manufacturer and have close partnerships with industry leading software and hardware vendors. All of our solutions are designed with scalability, longevity and the lowest total cost of ownership in mind. We do not use any proprietary hardware.
Find out more..Product
NVIDIA DGX1 certified reseller
Novatech are now a certified NVIDIA DGX-1 Partner, allowing us to supply our customers with the world's first purpose-built system for deep learning.
Read moreTrusted by
